Tesla’s Full Self-Driving system has made its long-awaited debut on Chinese roads, and initial reviews have been mixed to say the least. It appears the system is far from being able to handle the complex nature of Chinese traffic, leading to a series of disturbing incidents.
A particularly telling test, documented by a prominent Chinese automotive blogger and widely shared on social media, paints a picture of the system struggling to interpret basic road markings and traffic flow. The video footage shows the Tesla repeatedly and inexplicably entering lanes designated for bicycles, motorcycles, and scooters—a maneuver that is not only illegal but potentially dangerous. In one particularly disturbing instance, the system mistakes a bicycle lane for a right-turn lane and makes a sharp turn directly into the path of a stationary police car. The system also showed alarming disregard for dense lane markings, crossing them at an alarming rate, suggesting a fundamental misunderstanding of basic traffic rules.
Tesla’s strategy for developing FSD relies heavily on collecting and analyzing data from real-world driving conditions, an approach that has allowed for incremental improvements in North America. The expectation is that as more data is accumulated, the system will gradually adapt to the unique characteristics of Chinese roads. Early indications, however, are that this adaptation process will be far from smooth. A benchmark test conducted by a leading Chinese automotive industry review team, which pitted Tesla’s FSD against advanced driver assistance systems developed by local manufacturers Li Auto and Aito, revealed a significant difference in performance. A Tesla Model Y equipped with FSD racked up a staggering 34 traffic violations and required 24 driver interventions during the test. In contrast, the Li Auto L7 and Aito M9, while not without their limitations, demonstrated a significantly higher level of proficiency, registering fewer violations and requiring fewer interventions.
These early results raise serious questions about the readiness of Tesla’s FSD for the Chinese market. The system’s repeated misinterpretation of bike lanes, its disregard for dense markings, and its overall erratic behavior suggest a fundamental lack of understanding of the nuances of Chinese traffic patterns. While Tesla remains optimistic that the system will improve over time, this performance highlights the significant challenges involved in adapting autonomous driving technology to a diverse and dynamic driving environment. For now, it appears that Tesla’s FSD in China has a long and difficult road ahead, requiring significant refinement and adaptation before it can be considered a reliable and safe driving assistant.